Objective: Pumped storage power stations pump water to upper reservoirs when the electricity load is low; they release water to lower reservoirs to generate electricity when the electricity load is high. In these stations, cable tunnels, which connect underground transformers and ground switch stations, usually have high-fall, long-distance, and large-slope attributes. These affect the cable flame spread characteristics. Methods: This paper studies the effect of slope (0°-45°) on cable flame morphology, flame front distance, average flame spread rate, cable surface temperature, and ceiling temperature by conducting cable flame spreading experiments in a 1/10 small-scale cable tunnel. Thermocouples are used to measure the cable surface temperature and tunnel ceiling temperature. A camera is used to record the cable flame morphology, while a computer is used to record data from the camera and thermocouples. Results: The flame front distance and the average flame spread rate increase with the tunnel slope. When the tunnel slopes are 0°, 15°, 30°, and 45°, the cable flame front distances are 539, 783, 1076, and 1 300 mm, respectively, with average cable flame spread rates of 0.43, 0.92, 1.28, and 1.53 mm/s. With increasing slope, the first peak of the cable surface temperature moves upward, and so does the first peak of the tunnel ceiling temperature. When the slope is 0°, the first peak temperature and position of the cable surface and the ceiling are 650 ℃ and -0.25 m, respectively. When the slope is 15°, the first peak temperature and position of the cable surface and the ceiling are 400 ℃ and -0.50 m, respectively. When the slope is 30°, the cable surface's first peak temperature and position are 200℃ and -0.25 m, respectively, whereas those of the ceiling are 250 ℃ and -0.50 m, respectively. When the slope is 45°, the cable surface's first peak temperature and position are 200 ℃ and -1.00 m, respectively, whereas those of the ceiling are 250 ℃ and -1.50 m, respectively. The peak position of the tunnel ceiling temperature is farther than the peak position of the cable surface temperature. Conclusions: First, in the inclined cable tunnel, the Coandǎ and stacking effects increase the flame inclined angle and the preheating area of the unburned cable, with the unburned cable's heating rate and the average cable flame rate increasing. Second, the cable ceiling heating has little effect on cable flame spreading. However, the copper core inside the cable acts as a "heater" and a "radiator, " affecting the cable's burning behavior. The high-temperature core heats the unburned cable zone, increasing the preheating area and cable flame spread rate. Third, in the inclined tunnel, the stacking effect enhances the heat dissipation of cable burning. In the horizontal tunnel, the high-temperature ceiling heats the cable and increases the cable's burning time. Fourth, under the combined effects of longitudinal airflow inertia and thermal buoyancy forces, the peak position of the tunnel ceiling temperature is farther than the peak position of the cable surface temperature.